Amphosoma Baral

Index Fungorum number: IF 813478; Facesoffungi number: FoF14390

Type species: Amphosoma resinicola Baral & G. Marson

Etymology: referring to the presence of a spore body at both ends of ascospores.

Latin diagnosis: Apothecia rehydratata 0.1–0.5 mm diam., alba, roseo- aurantiaca vel atro-olivacea, sessilia vel breviter stipitata, superficialia vel immersa. Asci apice rotundati, tunica in statu emortuo plerumque incrassata, non amyloidei (IKI), octospori, in basi non uncinati, bifurcati. Ascosporae di- vel tristichae in ascis vivis, cylindrico-ellipsoideae vel fusoideo-clavatae, rectae, intus ad polis in statu vivo cum corpusculo lentiformi ad globoso, refringenti. Paraphyses ad apicem non vel valde inflatae, exsudato continuo vel granuloso tectae, cohaerentes vel non cohaerentes. Cellulae excipuli marginalis et paraphysium in statu vivo vacuolas globosas refringentes continentes. Habitat ad corticem, lignum et resinam, in ramis arborum plerumque coniferarum, in aere prominentibus.

Description: — TELEOMORPH: Apothecia rehydrated 0.1–0.5 mm diam., whitish to rose-orange or dark blackish-olivaceous, translucent when light-coloured, sessile to short-stalked, superficial to immersed. Asci *26–120 × 5.5–11 µm, protruding 1–3 µm beyond exudate when turgescent, 8-spored, spores 2–3-seriate inside living asci; apex hemispherical to conical or slightly truncate, with distinct apical thickening (rarely without); base simple septate, T-, L-, Y-, h- (or H-) shaped, exceptionally with true croziers with long slit-like perforation. Ascospores *(3.3–)5–15(–17.5) × (1.7–)2–4(–5) µm, cylindric-ellipsoid, also fusoid, usually homopolar in shape but some distinctly clavate (then some spores inversely oriented), straight, overmature sometimes becoming 1(–3)-septate, SBs *(0.2–)0.3–1(–1.2) × 0.5–1.3 µm, strongly refractive, lens- to plug-shaped or (sub)globose, ± broadly attached to wall at both spore ends. Paraphyses apically not or only slightly inflated, rarely strongly clavate-capitate or sublanceolate, *1.8–6 µm wide, terminal cells about as long as lower cells. Medullary excipulum of dense textura angularis(-intricata), indistinctly to medium sharply delimited. Ectal excipulum hyaline or with yellowish or olivaceous exudate, near base of t. angularis, at flanks and margin of t. angularis to t. prismatica oriented at a high or low angle, (†) slightly to medium gelatinized (common walls 1–1.5 µm thick), cells near base *5–12(–16) × 4–8(–11) µm; glassy processes and hairs absent. Anchoring hyphae *1.8–6 µm wide, wall 0.2–0.5 µm thick, sometimes covered by a thick gel sheath. SCBs not observed; VBs present in both paraphysis and ectal excipular cells (living state), scattered to abundant, consistently globose and strongly refractive, hyaline, 0.2–2 µm diam., staining bright turquoise in CRB, dead plasma showing a light copper-amber pigmentation due to oxidation of VBs; LBs with carotenoids sometimes present. Exudate 0.2–0.5 µm thick, continuous, externally smooth or warted to granular, hymenial elements conglutinated or ± easily separable by pressure, gel staining distinctly lilac in CRB. ― ANAMORPH: trinacrium-like. Colonies whitish to pale salmon-coloured, slow growing (CMA:2, MEA 2 %; A. resinicola). Conidiophores not observed. Conidia 2-armed (T-shaped), total size */†16–41.5 × 11–44 µm, stipe 1–6-septate, arms bent downwards, 0–5-septate.

Habitat: on dead, ± undecayed bark and wood of coniferous branches (usually associated with blackened resin), rarely on angiosperm bark. Desiccation tolerance: mature asci tolerating ~ 0.5–2 years. Distribution: Europe, North America, South Africa, (oro)temperate to (sub)tropical, humid to semi- or subhumid.

Recognized species: A. atroolivaceum, A. macrosporum, A. natalense, A. resinicola, A. aff. resinicola.

Taxonomy and relationship. The genus Amphosoma forms a small, undoubtedly natural group of species with desiccation- tolerant apothecia, which are predominantly conifer-inhabiting and associated with slightly decayed, blackened resinous wounds. Two characteristics are only visible in the living state, but serve as consistent key characteristics of the genus: (1) small, lens- to plug-shaped SBs occur at both ends of the ascospores; (2) refractive, globose VBs are regularly found inside the terminal cells of paraphyses and marginal ectal excipulum. When only dead specimens are studied, the genus Amphosoma may easily be misclassified in the Helotiales.

The bipolar arrangement of SBs and the preference for gymnosperms appear to argue for a phylogenetically old group within the Orbiliomycetes (see Phylogeny). Amphosoma seems to represent a connecting link between Hyalorbilia and Orbilia by sharing characteristics of both genera. Amphosoma resembles Hyalorbilia mainly in sporological features: the ascospores are more or less homopolar in shape, and the SBs bipolar- symmetrically arranged. For instance, Hyalorbilia subfusispora closely resembles Amphosoma in spore shape and bipolar SBs. Also the thin continuous exudate over the paraphyses of A. resinicola is reminiscent of a Hyalorbilia. However, Hyalorbilia differs in a number of traits, especially concerning ascus apex and base, orientation of the ectal excipulum at flanks, length of terminal cells of paraphyses in relation to lower cells, and type of VBs (see also Tab. 56).

The ascospores in the large genus Orbilia possess, in contrast to Amphosoma, only 1 SB per spore (usually at the upper end, rarely ± laterally), and the spores are often ± heteropolar in shape. As a rare exception, a few aberrant spores of Orbilia may contain two SBs, one at each end, reminiscent of Amphosoma. What underlines proximity of Amphosoma to Orbilia is that their asci have a bifurcate, generally simple- septate base and a more or less pronounced apical dome, sometimes with an apical chamber, reminiscent of subgenus Hemiorbilia. Amphosoma appears to represent the common ancestor of the large genus Orbilia, when assuming that in this genus the bifurcate ascus base lacking croziers evolved for the first time within the Orbiliomycetes.

The hymenial elements of Amphosoma are usually easily separable because the gel on the walls forms an only thin layer, which can be visualized by a deep lilac stain in CRB. The living paraphyses and marginal excipular cells contain strongly refractive, hyaline, globose VBs. This striking guttular pattern provides a characteristic being never observed in Hyalorbilia or Orbilia. The VBs are consistently globose and usually show no or only slight Brownian motion, apparently because the vacuoles in which they occur are rather small. However, in A. macrosporum strong Brownian motion was noted. In Hyalorbilia a similar guttular pattern inside the paraphyses and cortical cells may be seen in submature apothecia, but the VBs are larger and only slightly to medium refractive, sometimes pale chlorinaceous, and at progressed maturity they tend to fuse by forming large elongate bodies of a width corresponding to that of the cell lumen. No such fusion of VBs has been seen in Amphosoma.

The molecular closeness of Amphosoma and Retiarius (see below) raises the question whether it is justified to distinguish these two genera. The morphology of the teleomorph of Retiarius is unknown, and an adaptation to pollen as nutrient source was so far not determined in Amphosoma.

Species delimitation and confusion. The four recognized species are quite easily distinguished. Only in the rare case of albinism in A. atroolivaceum, delimitation against the shorter- spored A. resinicola was difficult.

Amphosoma persooniae Crous was proposed by Crous (2017) for a pycnidial fungus on stem canker of Persoonia sp. from southeastern Australia. It was placed in Amphosoma solely based on molecular data which clearly point to that genus (see below). We have doubts, however, that the described fungus is orbiliaceous. The species formed minute, globose, dark brown pycnidia which contain hyaline phialidic conidiogenous cells with prominent periclinal thickening that form 4.5–7 × 2–3 µm large, hyaline, smooth, subcylindrical conidia. For the time being we consider Amphosoma persooniae as non-orbiliaceous and the associated molecular data as belonging to an unknown species of Amphosoma. Besides, the taxon is invalid because the generic name was only proposed as a nomen provisorium in Baral et al. (2017b) (ICN Art. 35.1).

Anamorph. In three of the Amphosoma species a trinacrium- like anamorph was recorded in which the arms of the slender, Y- or T-shaped conidia are consistently bent downwards. This anamorph strongly resembles that found in Hyalorbilia erythrostigma and H. orbiliicola. Similar but more robust conidia with downwards curved arms resemble Trinacrium incurvum and occur in Orbilia in series Lentiformes, Piliferae, and Hesperideae (O. corculispora).

Phylogeny. Molecular data were available for three species, A. atroolivaceum, A. resinicola, A. aff. resinicola (G.M. 2015- 09-15.2), and the dubious A. persooniae, comprising ITS and LSU. A deviating sequence of A. atroolivaceum (G.M. 2017-11-05.1) includes also SSU V6–V9. The S1506 intron is absent in all of them. Our phylogenetic analyses of 5.8S+LSU or SSU+ITS+LSU (Phyls 1–2) support the position of Amphosoma near Lecophagus, Retiarius, and Lilapila by showing a higher distance to the morphologically similar genus Hyalorbilia. However, the distance to all these genera is very high, and even between A. atroolivaceum and A. resinicola it is 21 % in the ITS. In the combined analysis of SSU+ITS+LSU in Baral et al. (2017b), the genus appeared paraphyletic by forming together with Lecophagus and Lilapila a strongly supported clade, distant from the Hyalorbilia clade and Orbilia. In the present combined analysis of all orbiliaceous genera with DNA data (Phyl. 2), Amphosoma appeared as an unsupported monophyletic clade which formed a strongly supported clade only under inclusion of Retiarius (and Lecophagus). This result is astonishing, considering the high morphological similarity between Amphosoma and Hyalorbilia, particularly regarding the bipolar SB arrangement in the ascospores.

When analysing ITS (Phyl. 3), Amphosoma formed a strongly supported monophyletic clade, whereas when analysing LSU (Phyl. 4) its species clustered unresolved among the other genera. Based on these analyses we feel affirmed in recognizing Amphosoma and Retiarius as two different genera, which is supported by a number of distinctive nucleotides in the ITS and LSU (Tab. 58). In Magyar et al.’s (2017a, b) Bayesian analyses of ITS+LSU the two genera formed strongly supported monophyletic clades in an unsupported sister position. The two genera include also various environmental sequences of different taxa from GenBank (A1–A6, R1–R4 in Phyl. 3; A5– A6, R1–R4 in Magyar et al. l.c.). It seems interesting that in the SSU V8–V9 Amphosoma and Lecophagus concur at some informative positions (Pl. 56) at which other genera differ (no SSU was available for the genus Retiarius).

A. persooniae shows a 10.5 % ITS distance   to   A. aff. resinicola, 14.5 % to A. resinicola, and 19 % to A. atroolivaceum. In our analysis of SSU+ITS+LSU (Phyl. 2), it clustered strongly supported with A. resinicola in a clade, but when including A. aff. resinicola (not shown) it clustered strongly supported with this species (NJ) or unresolved with A. resinicola s.l. (ML). The morphological similarity between A. aff. resinicola and A. natalense, including a very similar spore size, might indicate that the A. persooniae sequence belongs to a teleomorph that closely resembles A. natalense. Both originate from angiosperm hosts and the present phylogeny suggests that angiosperm inhabiting species of Amphosoma developed from the so far predominant conifericolous taxa. The phylogeny of the angiosperm inhabiting sister genus

Retiarius appears to have taken place in a separate lineage.

Specific nucleotide positions. In the entire 5.8S region Am- phosoma atroolivaceum, A. aff. resinicola, and ‘A. persooniae’ concur with Retiarius revayae and R. superficiaris, whereas A. resinicola differs by 1 nt and R. bovicornutus by 2 nt (3 nt be- tween the latter two). Because of this similarity, Amphosoma concurs with Retiarius in the region of the ‘Orbiliales-specific’ primers Orb5.8s1F and Orb5.8s3F (Smith & Jaffee 2009), except that A. resinicola deviates in Orb5.8s3F by concurring with Lecophagus (Tab. 18). In the primer Orb28s2R, A. atrooli- vaceum and ‘A. persooniae’ concur with R. bovicornutus and R. superficiaris, but also with Lecophagus, Lilapila p.p., and some Orbilia spp., whereas A. resinicola and R. revayae deviate by the same nucleotide, and A. aff. resinicola shows another var- iant (Tab. 19). In the LSU D3 domain, 2 positions are unique for Amphosoma (including the strains from Texas): at pos. 748 and 770 the genus has GACCGGTCCTG and GGATCTGAG, respectively, in contrast to all other Orbiliomycetes (including Retiarius) which have GACCGATCCTG (very rarely other variants) and GGATTTGAG. Another motif (pos. 726–730) is characteristic for Amphosoma (including the strains from Texas) and Retiarius revayae (AGGTGGGAGCCCGCAAG), in contrast to various other variants of this motif in the remaining Orbiliomycetes. However, A. aff. resinicola has here AGGTGGGAGCCCGTAAG similar as Bryorbilia sp. (B2, AGGTGGGAGCCCGTAG). Members of the Vermispora clade of Hyalorbilia have a similar motif here, deviating only at pos. 722: AGGTGAGAGCCCGCAAG.

Ecology. The two European members of Amphosoma occur in the same habitats as those of Lilapila, i. e., resinous, xeric, undecayed coniferous bark or sometimes wood in montane to subalpine, orotemperate to orosubmediterranean, humid (mixed) conifer forests. A. atroolivaceum also occurs in temperate colline regions of central Europe. The single record of the North American A. macrosporum was on xeric, rotten, non-resinous bark of Pinus in the cold-temperate subhumid southwest, and that from South African A. natalense on xeric, medium decayed bark of a legume tree in a subtropical humid (winter-dry) coastal forest.

Species

  • Amphosoma resinicola